Postnatal Depression in New Zealand:
Findings from the 2015 New Mothers’ Mental Health Survey
May 2016
ISBN: 978-1-927303-76-4
Prepared for the Health Promotion Agency by:
Zoe Deverick and Hayley Guiney
HEALTH PROMOTION AGENCY
PO Box 2142
Wellington 6140
New Zealand
www.hpa.org.nz
May 2016
CONTENTS
Executive summary
4
Introduction
4
Method
4
Results
4
Conclusion
5
Introduction
6
Method
6
Design overview
6
Respondents
6
Questionnaire
7
Procedure
7
Data analysis
7
Results
9
Respondent characteristics
9
PND prevalence: Univariate analysis
10
PND prevalence: Multivariate analysis
16
Social factors and life experiences
17
Help-seeking
22
Results Summary
25
Prevalence estimate
25
Social and psychological experiences
26
Help-seeking
27
Discussion
28
Key findings
28
Limitations
28
Conclusion
29
Acknowledgements
30
References
31
Postnatal depression in New Zealand
3
EXECUTIVE SUMMARY
INTRODUCTION
The purpose of this report is to add to current knowledge around postnatal depression (PND) in
New Zealand by providing an indication of PND prevalence as well as an overview of the social
and life experiences, as well as help-seeking knowledge and attitudes, of women who might be
experiencing PND. To this end, the current report uses data from the New Mothers’ Mental Health
Survey (NMMHS), a cross-sectional survey conducted between July and September 2015.
METHOD
The NMMHS was conducted online, with 805 women who had given birth during the previous two
years. Respondents were recruited via the Bounty New Zealand database.
Questionnaire
Respondents completed the Edinburgh Postnatal Depression Scale (EPDS; Cox, Holden, &
Sagovsky, 1987) to assess PND (referred to here as EPDS-PND), and provided information on
their socio-demographic characteristics. Respondents also completed measures to assess anxiety
and psychological distress, and were asked questions to assess their social and life experiences,
as well as their knowledge, attitudes and behaviour around help-seeking.
Data analysis
Proportions were calculated first to estimate the prevalence of EPDS-PND in New Zealand.
Logistic regression analyses were then used to assess whether EPDS-PND prevalence varied by
socio-demographic factors, and whether social experiences and help-seeking behaviours differed
for those who met the criteria for EPDS-PND compared with those who did not.
RESULTS
PND prevalence
Overall, 14% of respondents met the criteria for EPDS-PND. After adjusting for covariates, people
who identified as Asian and those who had a low household income (under $40,000) were more
likely to meet the criteria for EPNDS-PND.
Social factors and connectedness
Respondents who met the criteria for EPDS-PND were more likely to give responses that indicated
greater life difficulties, lower coping self-efficacy, lower social connectedness, more isolation, lower
family/whānau wellbeing, and lower life satisfaction.
Help-seeking
The majority of respondents believed they could identify depression in themselves or in a friend,
regardless of whether or not they themselves met the criteria for EPDS-PND.
4
Postnatal depression in New Zealand
Those with EPDS-PND and those without had similar knowledge about the places they could go to
seek help if they thought they might be experiencing depression. The main potential source of help
listed by respondents (regardless of their EPDS-PND status) was their doctor or general
practitioner.
However, when asked where they would be most likely go for help if they thought they might be
experiencing depression, respondents who met the criteria for EPDS-PND were more likely to say
they did not know where they would go, and more likely to say they would not seek help.
CONCLUSION
This report provides useful insight into some of the characteristics and experiences of women
potentially experiencing PND in New Zealand.
Postnatal depression in New Zealand
5
INTRODUCTION
Postnatal depression encompasses symptoms of physical, emotional and psychological upheaval
experienced by women during the postpartum period. For many parents, the arrival of a baby may
be a time filled with both positive emotions and significant stressors due to the physical and
emotional demands of caring for a newborn. However, some women also experience significant
distress or disability due to symptoms of depression and anxiety experienced during this time. For
some, these symptoms may continue for up to a year or more postpartum (Bewley, 1999;
Goodman, 2004; Leahy-Warren & McCarthy, 2007).
There has been limited research into PND in a New Zealand context, but the few studies on PND
in New Zealand report prevalence rates typically ranging from 11 to 16% (Abbott & Williams, 2006;
Ekeroma et al., 2012; Thio et al., 2006). The purpose of this report is to add to current knowledge
around PND in New Zealand by using data from the 2015 New Mothers’ Mental Health Survey
(NMMHS). This report provides an indication of PND prevalence as well as an overview of the
social and life experiences, and help-seeking knowledge and attitudes, of women who might be
experiencing PND.
METHOD
DESIGN OVERVIEW
The NMMHS was an online survey designed to assess different aspects of mental health and
wellbeing among women who had given birth within the last two years. Due to the low population
base rate of new mothers, names and contact details of potential respondents were obtained from
an online panel (Bounty New Zealand). Note that the NMMHS was not designed to be nationallyrepresentative of all women who gave birth in the last two years, however it did utilise a large and
diverse sample in order to gain some insight into PND in the New Zealand context. Fieldwork was
conducted by National Research Bureau Ltd. (NRB), between July and September 2015.
RESPONDENTS
The names and contact details of 2,407 names of women who had recently given birth at locations
around New Zealand were provided by Bounty New Zealand. Of these, 1,770 were contacted by
the fieldwork company by telephone and invited to participate in the online survey.
Criteria for participation in the NMMHS were that respondents had given birth within the last two
years, and had provided Bounty New Zealand with a name, a telephone number, and an email
address. There were no further restrictions for participation. Respondents were placed in the draw
to win either an iPad Air or a Samsung Galaxy Tab 2 for completing the survey.
Two hundred and twenty women were initially contacted by telephone and invited to participate in a
pilot survey; 96 completed it. Following successful completion of the pilot, new respondents were
6
Postnatal depression in New Zealand
invited to participate in the final survey. The response rate was 56.5%, with a final sample size of
805 women.
QUESTIONNAIRE
The NMMHS was developed by HPA in consultation with the fieldwork company. The Edinburgh
Postnatal Depression Scale (EPDS), which was used as a measure of PND, was used with
appropriate permission. The survey also included questions to assess respondents’ sociodemographic characteristics, as well as their social and life experiences.
Measure of PND: EPDS
The EPDS is a validated screening tool designed to assess symptoms of depression within the
postpartum period, as distinct from the normative symptoms of pregnancy and postpartum
adjustment (Lussier, David, Saucier, & Borgeat, 1996). The EPDS comprises 10 items designed to
rate symptoms of PND, as experienced in the past seven days. Answers are on a four-point Likert
scale ranging from one (most of the time) to four (no, not at all). Total scores range from zero to
30, with scores of 13 or above considered indicative of EPDS-PND (Cox et al., 1987).
Other measures
This report also covers questions included in the NMMHS relating to respondents’ sociodemographic characteristics, feelings of social connectedness, family/whānau wellbeing (derived
with permission from Statistics New Zealand’s Te Kupenga 2013 questionnaire on Māori
wellbeing), personal wellbeing, and help-seeking attitudes.
PROCEDURE
To assess suitability of the survey questionnaire, a pilot study was first conducted with 96 new
mothers. Results from the pilot survey indicated that the questionnaire was performing to purpose,
and only minor wording changes were made to the final version.
Potential respondents for the final survey were contacted by telephone and invited to participate.
They were then sent an email link to the online questionnaire. Once respondents logged in to the
questionnaire, they were able to save their progress and return later. Email reminders were issued
to those who agreed to participate but did not commence, or only partially completed, the survey.
DATA ANALYSIS
Respondents were split into two groups, based on their EPDS scores. In line with
recommendations by Cox et al. (1987), respondents scoring 13 and above on the EPDS are
described in this report as meeting the criteria for EPDS-PND. Respondents scoring 12 or below
are described as not meeting the criteria for EPDS-PND, or having ‘no depression’. However,
because the EPDS is intended as a screening tool only, these terms should be interpreted as
indications of PND symptoms, rather than definitive diagnoses. Odds ratios (ORs), 95% confidence
intervals (CIs), and statistically significant p-values are reported for each comparison.
Postnatal depression in New Zealand
7
Prevalence analyses
EPDS-PND prevalence rates were examined by ethnicity, education, household income, mother’s
age, number of children, and age of youngest child. Initially, each socio-demographic variable was
examined using univariate logistic regression in relation to PND prevalence. All socio-demographic
variables in the univariate analysis were then considered together using multivariate logistic
regression, in order to control for confounding variables.
Social and life experiences and help-seeking analyses
Social experiences and help-seeking behaviours were assessed with questions relating to feelings
of social connectedness, family/whānau wellbeing, help-seeking attitudes, personal wellbeing, and
psychological experiences. For questions that used a five-point agreement scale, “agree” and
“strongly agree” responses were combined to indicate total agreement. When respondents
selected “don’t know” or “prefer not to say”, that questionnaire item was not scored. Logistic
regression was then used to compare differences in the social experiences and help-seeking
behaviour of respondents who met the criteria for EPDS-PND to those who did not.
8
Postnatal depression in New Zealand
RESULTS
RESPONDENT CHARACTERISTICS
A total of 805 new mothers completed the survey, the socio-demographic characteristics of whom
are shown in Table 1. Note that the number of respondents in each category may not add to the
total sample because some respondents did not answer every question.
Table 1. Demographic characteristics of all respondents
N
Overall
%
805
Ethnicity
Māori
97
12
Pacific
29
4
Asian
71
9
NZ European/Other
606
75
No formal
38
5
Secondary
239
31
Trade certificate/professional qualification
174
23
Degree/postgrad
319
41
<20
49
7
20-40
71
10
40-60
156
22
60-80
135
19
80-100
135
19
100-150
124
17
>150
55
8
0-1
745
93
1-2
54
7
15-24
104
13
25-34
490
61
>35
211
26
1
462
57
2
231
29
3
70
9
4+
42
5
Education
Total household income ($1,000 over the past 12 months)
Youngest child age
Mother’s age
Number of children
Postnatal depression in New Zealand
9
PND PREVALENCE: UNIVARIATE ANALYSIS
EPDS scores were calculated for 769 of respondents who completed each item in the EPDS.
Scores ranged from zero to 25, with a mean score of 7.2.
Overall prevalence of EPDS-PND amongst all respondents was consistent with rates previously
reported in New Zealand, in that 14% (95% CI [11, 16%]) of all respondents met the criteria for
EPDS-PND (Abbott & Williams, 2006; Carter et al., 2005; Ekeroma et al., 2012; Gao et al., 2007,
2010; McGill et al., 1995; Thio et al., 2006; Waldie et al., 2015; Webster et al., 1994).
To assess the rate of PND in different socio-demographic subgroups, differences in the percentage
of respondents who met the criteria for EPDS-PND in each subgroup were examined using
univariate logistic regression. The socio-demographic variables examined were ethnicity, highest
educational qualification, household income, mother’s age, number of children, and age of
youngest child.
10
Postnatal depression in New Zealand
Ethnicity
All respondents were asked which ethnic group they belonged to. Respondents could choose more
than one ethnicity, however when multiple selections occurred, ethnicity was prioritised in the order
of Māori, Pacific, Asian, and NZ European/Other, as recommended by the Ethnicity Data Protocols
for the Health and Disability Sector (Ministry of Health, 2004).
Percentages of respondents in each ethnic group who met the criteria for EPDS-PND are
displayed in Figure 1. Twenty three percent (95% CI [14, 35%]) of those who identified as Asian
met the criteria for EPDS-PND, compared with 13% (95% CI [10, 16%]) of those who identified as
NZ European or Other, 12% (95% CI [7, 21%]) of those who identified as Māori, and 11% (95% CI
[4, 30%]) of those who identified as Pacific. However, the prevalence rate for respondents who
identified as Pacific may be an unreliable estimate due to a small sample size for this group.
Univariate logistic regression indicated a significant effect of Asian ethnicity, in that respondents
who identified as Asian were more likely than those who identified as NZ European or Other to
meet the criteria for EPDS-PND (OR = 2.05, 95% CI [1.10, 3.83], p = .03). There were no
significant differences in prevalence rates for respondents identifying with other ethnicities,
although note that previous research with a larger sample size has indicated higher rates of EPDSPND among Pacific women (Abbott & Williams, 2006).
Figure 1. Percentages of respondents who met the criteria for EPDS-PND, by ethnicity
100%
90%
Percentage EPDS-PND
80%
70%
60%
50%
40%
30%
23%
12%
13%
20%
10%
0%
Maori
Asian
Ethnicity
Euro/Other
Note: Respondents of Pacific Island ethnicity are not shown as the estimate was unreliable (due to
insufficient sample size).
Postnatal depression in New Zealand
11
Education
All respondents were asked to indicate their highest educational qualification.
Percentages of respondents who met the criteria for EPDS-PND are displayed by highest
educational qualification in Figure 2. Sixteen percent (95% CI [7, 32%]) of those who reported
having no formal education met the criteria for EPDS-PND; however, this estimate may be
unreliable due to a small sample size for this group. Seventeen percent (95% CI [13, 22%]) of
respondents who selected secondary school as their highest qualification met the criteria for
EPDS-PND, compared with 15% (95% CI [11, 22%]) who selected “trade certificate or professional
qualification”, and 10% (95% CI [7, 14%]) who selected “bachelor’s degree or higher”.
Univariate logistic regression indicated that PND prevalence varied with education level, in that
respondents who reported having higher educational qualifications were less likely to meet the
criteria for EPDS-PND. Specifically, those whose highest qualification was secondary school
completion were more likely than those with at least a bachelor’s degree to meet the criteria for
EPDS-PND (OR = 1.90, 95% CI [1.14, 3.17] p = .02).
Figure 2. Percentages of respondents who met the criteria for EPDS-PND, by education
level
100%
Percentage EPDS-PND
90%
80%
70%
60%
50%
40%
16%
30%
17%
20%
15%
10%
10%
0%
No formal
12
Secondary
Trade cert/prof
qualification
Highest qualification
Bachelor's degree or
higher
Postnatal depression in New Zealand
Household income
All respondents were asked which income bracket best fit their total household income before tax,
over the last 12 months.
Percentages of respondents who met the criteria for EPDS-PND in each income bracket are
displayed in Figure 3. Univariate logistic regression showed that EPDS-PND prevalence varied by
household income, such that for every step increase in income bracket, the odds of meeting
criteria for EPDS-PND decreased by 0.73 (OR = 0.73, 95% CI [0.64, 0.84] p < .001).
Figure 3. Percentages of respondents who met the criteria for EPDS-PND by income bracket
100%
90%
Percentage EPDS-PND
80%
70%
60%
50%
29%
40%
22%
30%
15%
20%
11%
8%
10%
4%
10%
0%
<20
20-40
Postnatal depression in New Zealand
40-60
60-80
80-100
Household income bracket
100-150
>150
13
Mother’s age
Respondents were asked to indicate which age bracket they belonged to.
Percentages of those who met the criteria for EPDS-PND in each age category are displayed in
Figure 4. Twenty three percent (95% CI [16, 32%]) of those in the 15 to 24 age category met the
criteria for EPDS-PND, compared with 13% (95% CI [10, 16%]) of those in the 25 to 34 age
category, and 11% (95% CI [7, 16%]) of those in the over 35 age category.
Univariate logistic regression indicated EPDS-PND prevalence varied by mother’s age, in that
younger respondents (in the 15-24 age category) were more likely than older respondents (in the
over 35 age category) to meet the criteria for EPDS-PND (OR = 2.45, 95% CI [1.23, 4.68], p = .01).
Figure 4. Percentages of respondents who met the criteria for EPDS-PND, by age category
100%
90%
Percentage EPDS-PND
80%
70%
60%
50%
40%
23%
30%
20%
13%
11%
25-34
Mother's age in years
>35
10%
0%
15-24
14
Postnatal depression in New Zealand
Number of children
Respondents were asked to indicate how many children they had who were under 18 years of age.
Categories were one, two, three, and four or more children.
Univariate logistic regression indicated no significant effects of number of children on the
prevalence of EPDS-PND.
Age of youngest child
To give an estimate of time postpartum, respondents were asked about the ages, in years, of each
of their children. This information was used to calculate the age of their youngest child.
The percentage of those who met the criteria for EPDS-PND are displayed by age of youngest
child in Figure 6. Thirteen percent (95% CI [11, 15%]) of women whose youngest child was under
one-year-of-age met the criteria for EPDS-PND, compared with 22% (95% CI [13, 35%]) of those
whose youngest child was aged between one and two years.
Univariate logistic regression indicated that respondents whose youngest child was aged between
one and two years were only marginally (but not significantly) more likely than respondents whose
youngest child was under one-year-of-age to meet the criteria for EPDS-PND (OR = 1.94, 95% CI
[0.99, 3.83], p = .06).
Figure 5. Percentage of respondents who met the criteria for EPDS-PND, by age of youngest
child
100%
90%
Percentage EPDS-PND
80%
70%
60%
50%
40%
22%
30%
20%
13%
10%
0%
0-1
1-2
Age of youngest child (years)
Postnatal depression in New Zealand
15
PND PREVALENCE: MULTIVARIATE ANALYSIS
Results from the initial univariate analyses indicated that PND prevalence varied by respondents’
ethnicity, household income, education level, and mother’s age. Because several of these
characteristics are likely to be covariates (eg, income may covary with education level and age),
socio-demographic characteristics and PND prevalence were placed into a multivariate model to
control for the relative effects of each variable.
After all other significant socio-demographic variables were controlled for, Asian ethnicity
(compared with NZ European/Other; p = .05) and lower household income (p = .04) emerged as
the only significant predictors of EPDS-PND. Results from the multivariate analysis with adjusted
odds ratios (AOR) 95% confidence intervals (CI) are displayed in Table 2, with significant variables
highlighted in bold.
Table 2. Multivariate model results
Adjusted Odds Ratio
(95% CI)
p-value
Māori
0.55 (0.24, 1.26)
0.14
Pacific
0.35 (0.08, 1.64)
0.19
Asian
2.09 (1.02, 4.23)
0.04
No formal
1.32 (0.43, 4.05)
0.62
Secondary school
1.35 (0.73, 2.49)
0.34
Trade cert
1.17 (0.61, 2.24)
0.63
0.74 (0.63, 0.88)
<.001
15-24
1.65 (0.76, 3.60)
0.21
25-34
1.06 (0.60, 1.88)
0.84
Ethnicity (ref = NZ European/Other)
Education (ref = degree/postgrad)
Income
Overall effect
Mother’s age (ref = 35+)
16
Postnatal depression in New Zealand
SOCIAL FACTORS AND LIFE EXPERIENCES
Respondents were asked to indicate their agreement with statements relating to perceived life
difficulty, social support and connectedness, feelings of isolation, and family wellbeing. Responses
from those who met the criteria for EPDS-PND were compared with those who did not meet the
criteria. On every item relating to social factors and connectedness, those who met the criteria for
EPDS-PND were less likely to agree with positive statements relating to these areas, and more
likely to agree with statements that indicated difficulties in these areas.
Life difficulty and coping self-efficacy
To assess perceived life difficulty and coping self-efficacy, respondents were asked to indicate
their agreement with the statements, “The last 12 months have been among the most difficult times
of my life”, and “I am able to cope with everyday stresses of life.”
Overall, one-half (51%; 95% CI [48, 55%]) of respondents agreed that the last 12 months had been
the most difficult of their life, while three quarters (77%; 95% CI [74, 80%]) agreed they were able
to cope with everyday stresses. Respondents who met the criteria for EPDS-PND were less likely
than those who did not meet the criteria to agree with either statement (Table 3).
Table 3 shows the percentage of respondents in each group who agreed with statements relating
to life difficulty and coping self-efficacy. Odds ratios, 95% confidence intervals, and p-values are
shown for respondents who met the criteria for EPDS-PND, compared with those who did not.
Table 3. Percentage of respondents endorsing questions relating to life difficulty
No
depression
EPDS-PND
(ref)
OR
(95% CI)
p-value
“The last 12 months have been among the
most difficult times of my life.”
(Strongly agree/agree)
47%
79%
4.24
(2.58, 6.96)
<.001
“I am able to cope with everyday stresses of
life.” (Strongly agree/agree)
85%
31%
0.08
(0.05, 0.12)
<.001
Item
Postnatal depression in New Zealand
17
Social support and connectedness
To assess perceived social connectedness, respondents were asked to indicate their agreement
with the statements, “I make an effort to see family or friends I don’t live with”, and “I can always
rely on a friend or family member for support if I need it.” Overall, 83% (95% CI [80, 85%]) of
respondents agreed that they made an effort to see friends and family. A similar proportion (85%;
95% CI [83, 88%]) agreed that they could always rely on a friend or family member for support.
Respondents who met the criteria for EPDS-PND were less likely than those who did not meet the
criteria to agree with either of these statements (see Table 4).
Respondents were also asked to indicate how easy (five-point scale: very easy to very hard) it
would be for them to find someone to (i) help with things such as pets, mail, or checking their
house, if they were away from home; and (ii) help them in times of need, for example providing a
place to stay if they suddenly needed it. Overall, 76% (95% CI [73, 79%]) of respondents indicated
that it would be “easy” or “very easy” for them to find someone to help if they were away from
home; a similar proportion (78%; 95% CI [75, 80%]) indicated that it would be either “easy” or “very
easy” for them to find someone to help them with a place to stay if they suddenly needed one.
Respondents who met the criteria for EPDS-PND were less likely than those who did not meet the
criteria respond “easy” or “very easy” to either of these statements (Table 4).
Table 4 shows the percentage of respondents in each group who responded positively to items
relating to social support and connectedness. ORs, 95% CIs, and p-values are shown for
respondents who met the criteria for EPDS-PND, compared with those who did not.
Table 4. Percentage of respondents who endorsed questions relating to social support and
connectedness
Item
No
depression
EPDSPND (ref)
OR
(95% CI)
p-value
“I make an effort to see family or friends I don’t
live with.” (Strongly agree/agree)
86%
67%
0.34
(0.21, 0.53)
<.001
“I can always rely on a friend or family member
for support if I need it”
(Strongly agree/agree)
90%
61%
0.17
(0.11, 0.29)
<.001
“How easy or hard would it be to find someone
to help you with things such as looking after
pets, collecting mail, or checking your house
while you are away from home?”
(Very easy/easy)
80%
56%
0.31
(0.20, 0.48)
<.001
“How easy would it be to find someone to help
you in times of need, such as providing a place
to stay if you suddenly needed one?”
(Very easy/easy)
82%
56%
0.28
(0.18, 0.43)
<.001
18
Postnatal depression in New Zealand
Feelings of isolation
To assess feelings of isolation, respondents were asked to indicate how often in the last four
weeks (five-point scale: none of the time to all of the time) they had felt isolated from others.
Overall, 67% (95% CI [63, 70%]) said that they had felt isolated at least some of the time.
Respondents who met the criteria for EPDS-PND were more likely than those who did not meet the
criteria to say that they had felt isolated at least some of the time (Table 5).
Respondents were also asked to indicate whether or not (yes/no) they had felt personally excluded
from a social situation over the past year. Overall, 43% (95% CI [40, 47%]) said that there had
been such an occasion. Respondents who met the criteria for EPDS-PND were more likely than
those who did not meet the criteria to respond “yes” to this item (see Table 5).
Table 5 shows the percentage of respondents in each group who endorsed questions relating to
isolation and social exclusion. ORs, 95% CIs, and p-values are shown for respondents who met
the criteria for EPDS-PND, compared with those who did not.
Table 5. Percentages of respondents who endorsed questions relating to feelings of
isolation
No
EPDS-PND
Item
OR (95% CI)
p-value
depression
(ref)
“In the last four weeks, how often have
you felt isolated from others?”
(All/most/some/a little of the time)
62%
98%
31.88
(7.80, 130.32)
<.001
“In the past year, has there been an
occasion when you felt personally
excluded from a social situation?” (Yes)
38%
73%
4.40
(2.75, 7.04)
<.001
Postnatal depression in New Zealand
19
General family/whānau wellbeing
To assess general family/whānau wellbeing, respondents were asked to indicate how well their
family/whānau (i) got along with one another and (ii) was doing these days. Response options
ranged from one (very well) to five (very badly). Overall, 83% (95% CI [85, 89%]) of respondents
indicated that their family/whānau got along well or very well; a similar proportion (86%; 95% CI
[83, 88%]) indicated their family/whānau was currently doing well or very well. Respondents who
met the criteria for EPDS-PND were less likely than those who did not meet the criteria to respond
‘well’ or ‘very well’ to either of these questions (Table 6).
Table 6 shows the percentage of respondents in each group who responded ‘well’ or ‘very well’ to
items relating to general family/whānau wellbeing. ORs, 95% CIs, and p-values are shown for
respondents who met the criteria for EPDS-PND, compared with those who did not.
Table 6. Percentages of respondents who endorsed questions relating to family/whānau
wellbeing
EPDS-PND
OR
Item
No depression
p-value
(ref)
(95% CI)
“In general, how would you rate the
way your family/whānau get along
with one another?” (Very well/well)
91%
68%
0.20
(0.12, 0.33)
<.001
“In general, how would you rate how
your family/whānau is doing these
days?” (Very well/well)
90%
64%
0.21
(0.13, 0.33)
<.001
20
Postnatal depression in New Zealand
Personal wellbeing
To assess perceived life satisfaction, respondents were asked “Overall, how satisfied are you with
life as a whole these days?” Response options ranged from one (very satisfied) to five (very
dissatisfied). Overall, 80% (95% CI [77, 83%]) indicated that they were satisfied, or very satisfied,
with their life as a whole. Respondents who met the criteria for EPDS-PND were less likely than
those who did not meet the criteria to say that they were satisfied or very satisfied (see Table 7).
Respondents were also asked to indicate the extent to which they felt that the things they did in
their life were worthwhile. Response options ranged from one (very worthwhile), to five (not at all
worthwhile). Overall, 83% (95% CI [81, 86%]) indicated that they felt the things they did in their life
were worthwhile or very worthwhile. Respondents who met the criteria for EPDS-PND were less
likely than those who did not meet the criteria to say that they felt the things they did in their life
were worthwhile or very worthwhile (see Table 7).
Table 7 shows the percentage of respondents in each group who responded positively to items
relating to personal wellbeing. ORs, 95% CIs, and p-values are shown for respondents who met
the criteria for EPDS-PND, compared with those who did not.
Table 7. Percentages of respondents who endorsed questions relating to personal
wellbeing
EPDS-PND
OR
Item
No depression
p-value
(ref)
(95% CI)
“Overall, how satisfied are you with
life as a whole these days?”
(Very satisfied/satisfied)
87%
44%
0.11
(0.07, 0.18)
<.001
“Overall, to what extent do you feel
the things you do in your life are
worthwhile?”
(Very worthwhile/worthwhile)
90%
49%
0.11
(0.07, 0.18)
<.001
Postnatal depression in New Zealand
21
HELP-SEEKING
Recognising depression
Respondents were asked to indicate their level of agreement with the items “If a friend had
depression, you would be able to recognise the signs and symptoms”, and “You would be able to
recognise if you had depression”. Most respondents agreed they would be able to recognise
depression in a friend (71%; 95% CI [67, 74%]) or in themselves (77%; 95% CI [74, 80%]); there
were no differences between those who met the criteria for EPDS-PND and those who did not.
Knowing where to seek help
To assess respondent’s level of knowledge of where they could seek help for depression, the item
“If you or someone you know has depression, do you know where you could go to get help?” had
an open answer format, with multiple answers accepted. The majority (71%; 95% CI [68, 74%])
listed their doctor or general practitioner; 15% (95% CI [12, 17%]) did not provide an answer to this
question, indicating either a refusal to answer or a lack of knowledge. Figure 6 shows potential
sources of help reported by at least 5% of respondents.
Figure 6. Respondent’s answers to the question, “If you or someone you know has
depression, do you know where you could get to get help?”
100%
90%
Percentage overall
80%
71%
70%
60%
50%
40%
30%
19%
19%
19%
20%
16%
15%
7%
10%
0%
Sources of help
Note: Places listed by fewer than 5% of respondents are not shown
There were no between-group differences in the proportion of respondents citing particular sources
of help, except that respondents who met the criteria for EPDS-PND (9%; 95% CI [5, 9%]) were
less likely than those who did not meet the criteria (17%; 95% CI [14, 20%]) to list a helpline (OR =
.47, 95% CI [0.23, 0.95], p = .04).
22
Postnatal depression in New Zealand
Help-seeking intentions
To assess intended help-seeking behaviour, respondents were asked “If you thought you might be
experiencing depression, do you know who, or where, would you be most likely to turn for help?”
Respondents could select from the options “yes”, “no”, or “nobody/nowhere/would not seek help”.
Overall, 82% (95% CI [79, 84%]) indicated they knew where they would most likely turn for help.
Thirteen percent (95% CI [11, 16%]) indicated they did not know where they would go, and 5%
(95% CI [4, 7%]) indicated they would not seek help.
Respondents who met the criteria for EPDS-PND were less likely than those who did not to say
they knew where they would go for help (OR = .38, CI [0.24, 0.60], p < .001) and more likely to say
that they would not seek help (OR = 5.57, CI [2.80, 11.07], p <.001).
Figure 7 displays the percentage of respondents in each group in relation to their answer to this
item.
Figure 7. Percentages of respondents who met the criteria for EPDS-PND, and those who
did not, in response to the question, “If you thought you might be experiencing depression,
do you know who or where you would be most likely to turn for help?”
No depression
EPDS-PND
100%
90%
85%
68%
80%
Percentage
70%
60%
50%
40%
30%
16%
20%
15%
12%
10%
3%
0%
Yes
Postnatal depression in New Zealand
No
Would not
"Do you know where you would turn for help?"
23
Respondents who said they knew who or where they would be most likely to turn for help were
asked to provide one qualitative response to the item “Please say who or where would you be most
likely to turn to for help?” Although it was intended that respondents would provide one answer
only (ie, the ‘most likely’ help source), many respondents provided more than one answer.
Therefore, this question was treated as a multi-response item.
Overall percentages of answers reported by at least 5% of respondents are displayed in Figure 8.
The most common overall responses were the doctor, followed by family and friends, followed by
their partner.
There were no significant between-group differences in answers to this question.
Figure 8. Most common answers to the question, “Who or where would you be most likely
to turn to for help?”
100%
90%
80%
Percentage
70%
60%
50%
44%
39%
40%
26%
30%
20%
6%
10%
0%
Doctor
Partner
Family/friends
Therapist
"Where would you turn for help?"
Note: Places listed by fewer than 5% of respondents are not shown
24
Postnatal depression in New Zealand
RESULTS SUMMARY
PREVALENCE ESTIMATE
The initial analysis indicated that PND prevalence was higher in respondents with Asian ethnicity,
lower household income, lower levels of formal education, and younger age. After adjusting for
covariates, Asian ethnicity and low household income were the only significant predictors of EPDSPND.
Table 8 provides an overview of the association between discrete socio-demographic factors and
PND prevalence. Significant findings are in bold.
Table 8. Socio-demographic factors associated with PND prevalence, univariate and
multivariate analysis
Univariate
Multivariate
PND prevalence
OR
(95% CI)
Ethnicity
Asian > NZ
European/Other*
2.05
(1.10, 3.83)
.03
Education
Secondary school
> Degree*
1.90
(1.14, 3.17)
.02 Not sig.
Income
Prevalence
decreased as
incomes rose
0.73
(0.64, 0.84)
Mother’s
age
(years)
15-24 > 35+*
2.45
(1.23, 4.68)
p PND prevalence
Asian > NZ
European/Other*
Prevalence
<.001 decreased as
incomes rose
.01 Not sig.
OR
(95% CI)
p
2.08
(1.02, 4.25)
.04
1.35
(0.73, 2.49)
0.34
0.74
(0.63, 0.88)
.04
1.65
(0.76, 3.60)
0.21
*=reference group, ‘Not sig.’ = not significant.
Postnatal depression in New Zealand
25
SOCIAL AND PSYCHOLOGICAL EXPERIENCES
This report found that for every question relating to the social and psychological experiences of
new mothers, those meeting the criteria for EPDS-PND were more likely than those who did not
meet the criteria to answer in a way that reflected problems across these areas.
Table 9 provides an overview of the differences in social and personal experiences for those who
met the criteria for EPDS-PND, compared with those who did not.
Table 9. ORs, CIs, and p-values for responses by those who met the criteria for EPDS-PND
(reference group = respondents who did not meet the criteria).
Group most likely to answer
‘yes’, or endorse item.
OR (95% CI)
p-value
Last 12 months most difficult of
life
EPDS-PND > No depression
4.24 (2.59, 6.96)
<.001
Can cope with everyday stress
No depression > EPDS-PND
0.08 (0.05, 0.12)
<.001
Make effort to see friends and
family
No depression > EPDS-PND
0.34 (0.21, 0.53)
<.001
Can rely on friends/family for
support
No depression > EPDS-PND
0.17 (0.11, 0.29)
<.001
Easy to find someone to help
No depression > EPDS-PND
0.31 (0.20, 0.48)
<.001
Easy to find somewhere to stay
No depression > EPDS-PND
0.28 (0.18, 0.43)
<.001
Feel isolated
EPDS-PND > No depression
31.88 (7.80, 30.32)
<.001
Felt excluded
EPDS-PND > No depression
4.40 (2.75, 7.04)
<.001
Family/whānau gets along
No depression > EPDS-PND
0.20 (0.12, 0.33)
<.001
Family/whānau doing well
No depression > EPDS-PND
0.21 (0.13, 0.33)
<.001
Satisfied with life
No depression > EPDS-PND
0.11 (0.07, 0.18)
<.001
Life is worthwhile
No depression > EPDS-PND
0.11 (0.07, 0.18)
<.001
26
Postnatal depression in New Zealand
HELP-SEEKING
When asked where they could go to seek help for depression, most respondents listed their doctor
or general practitioner. Other places listed, in order of frequency, were: family and friends, a
therapist, the internet, a helpline, and Plunket. There were few differences between respondents
who met the criteria for EPDS-PND and those who did not in awareness of sources of help, other
than for ‘helpline’, which was less likely to be listed by respondents who met the criteria for EPDSPND than those who did not. However, when asked where they actually would go for help,
respondents who met the criteria for EPDS-PND were less likely than those who did not to say
they knew where they would go, and more likely to say that they would not seek help.
Table 10 provides an overview of significant comparisons between respondents who met the
criteria for EPDS-PND, compared with those who did not, on items relating to help-seeking.
Table 10. Significant comparisons between respondents who met the criteria for EPDS-PND,
and those who did not, on items relating to help-seeking.
Group comparison
OR (95% CI)
p-value
List helpline as place where could
seek help
No depression > EPDS-PND
0.47
(0.23, 0.95)
.04
Know where would seek help
No depression > EPDS-PND
0.38
(0.24, 0.60)
<.001
Would not seek help
EPDS-PND > No depression
5.57
(2.80, 11.07)
<.001
Postnatal depression in New Zealand
27
DISCUSSION
The purpose of this report was to use data from the NMMHS to provide an indication of PND
prevalence in New Zealand and to gain insight into some of the socio-demographic differences of
women who might be experiencing PND in New Zealand. It also sought to understand the
experiences of respondents who met the criteria for EPDS-PND in the areas of social
connectedness, family/whānau wellbeing, life difficulties, self-efficacy, personal wellbeing, and
help-seeking.
KEY FINDINGS
EPDS-PND prevalence
When confounding variables were controlled, Asian ethnicity and a low household income
emerged as the only variables significantly associated with higher rates of EPDS-PND. These
findings are consistent with prior New Zealand research conducted in New Zealand (Gao et al.,
2010; Waldie et al., 2015; Webster et al., 1994), although the higher PND rate amongst women of
Pacific Island ethnicity identified in previous research (Waldie et al., 2015) was not observed in the
current study, possibly due to the sample of Pacific Island women being too small to reliably infer
prevalence for this group.
Social and life experiences
When asked about their social and life experiences, respondents meeting the criteria for EPDSPND were more likely than those with no depression to have negative experiences in each of these
areas. For example, respondents with EPDS-PND reported less social support and
connectedness, less perceived coping self-efficacy, lower personal wellbeing, and more feelings of
social isolation and exclusion.
These findings indicate that women experiencing PND encounter compounding risk factors, such
as low general family wellbeing, and social isolation. Although the cross-sectional design of the
NMMHS means that it is not possible to infer whether low social support precipitates PND or the
other way around, it is important to note that previous research suggests the relationship is
bidirectional (Koenders et al., 2015).
Help-seeking
The majority of respondents were able to list places where they could seek help for depression and
there were no differences in awareness of potential help services between those with EPDS-PND
and those with no depression. However, respondents with EPDS-PND were less likely to say they
knew where they would actually go to seek help and more likely to say they would not seek help.
LIMITATIONS
Results from the current research must be considered in light of the following potential limitations.
While the NMMHS was designed to gain insight into PND in a large nationwide sample of new
mothers, the sample was not nationally-representative. This means that accurate prevalence rates
28
Postnatal depression in New Zealand
for Pacific Island respondents could not be reliably estimated, due to a low sample size in this
group.1 Further, the online nature of the survey means that some participants may have been
excluded if they did not have access to a computer and internet connection, particularly if they
were in lower socio-demographic categories. Note however, that the NMMHS included an
adequate sample of respondents from lower education and income brackets. Finally, the NMMHS
assessed respondents over a wide postnatal period: women were eligible to participate if they had
given birth at any point over the past two years. Because PND prevalence is proposed to peak at
between six and 12 weeks postpartum (Leahy-Warren & McCarthy, 2007), this wide time frame
may have captured women who might have had PND initially but since recovered or women who
were experiencing distress not related to the postpartum period.
CONCLUSION
This report provides useful insight into some of the characteristics and experiences of women in
New Zealand who might be experiencing PND. However, its cross-sectional nature means that it is
not possible to infer causality between PND and the socio-demographic characteristics and life
experiences of respondents. Indeed, the relationships between some of these variables are likely
to be bi-directional. Furthermore, the small number of respondents identifying as Pacific limited the
ability to understand PND in that population. To address these issues, future research could
consider implementing longitudinal designs with nationally-representative samples, to delineate the
relationship between socio-demographic characteristics, difficult life experiences, and PND.
1
For more information on postnatal depression in Pacific Island women living in New Zealand, readers are referred to prior New
Zealand research by Abbott and Williams (2006), which includes a large sample of Pacific Island mothers.
Postnatal depression in New Zealand
29
ACKNOWLEDGEMENTS
The authors would like to acknowledge the following parties for their generous and important
contributions to this research:
National Research Bureau Ltd for collecting the data used in the NMMHS.
Respondents who took the time to participate in this research.
People who contributed to the development of the NMMHS.
Scholars who gave permission to use their tools in this survey.
Statistics New Zealand for their permission to use family/whānau wellbeing questions from
Te Kupenga 2013.
Dr Karen McBride-Henry, Michelle Mako, and Virginia MacEwan for their peer reviews of
this report.
Postnatal depression in New Zealand
30
REFERENCES
Abbott, M. W., & Williams, M. M. (2006). Postnatal depressive symptoms among Pacific mothers in
Auckland: prevalence and risk factors. Australian & New Zealand Journal of Psychiatry,
40(3), 230–238.
Bewley, C. (1999). Postnatal depression. Nursing Standard, 13(16), 49–54; quiz 55–6.
Carter, F. A., Carter, J. D., Luty, S. E., Wilson, D. A., Frampton, C. M. A., & Joyce, P. R. (2005).
Screening and treatment for depression during pregnancy: a cautionary note. Australian &
New Zealand Journal of Psychiatry, 39(4), 255–261.
Cox, J. L., Holden, J. M., & Sagovsky, R. (1987). Detection of postnatal depression. Development
of the 10-item Edinburgh Postnatal Depression Scale. The British Journal of Psychiatry,
150(6), 782–786.
Gao, W., Paterson, J., Abbott, M., Carter, S., & Iusitini, L. (2007). Maternal mental health and child
behaviour problems at 2 years: findings from the Pacific Islands Families Study. Australian
& New Zealand Journal of Psychiatry, 41(11), 885–895.
Gao, W., Paterson, J., Abbott, M., Carter, S., & Iusitini, L. (2010). Pacific Islands Families Study:
Intimate Partner Violence and Postnatal Depression. Journal of Immigrant and Minority
Health, 12(2), 242–248. http://doi.org/10.1007/s10903-008-9190-y
Goodman, J. H. (2004). Postpartum depression beyond the early postpartum period. Journal of
Obstetric, Gynecologic, & Neonatal Nursing, 33(4), 410–420.
Koenders, M. A., Giltay, E. J., Hoencamp, E., Elzinga, B. M., Spinhoven, P., & Spijker, A. T.
(2015). The bidirectional impact of perceived and enacted support on mood in bipolar
outpatients: A two-year prospective study. Comprehensive Psychiatry, 60, 59–67.
http://doi.org/10.1016/j.comppsych.2015.03.009
Leahy-Warren, P., & McCarthy, G. (2007). Postnatal Depression: Prevalence, Mothers’
Perspectives, and Treatments. Archives of Psychiatric Nursing, 21(2), 91–100.
http://doi.org/10.1016/j.apnu.2006.10.006
Lussier, V., David, H., Saucier, J.F., & Borgeat, F. M. (1996). Self-rating assessment of postnatal
depression: A comparison of the Beck depression inventory and the Edinburgh postnatal
depression scale. Pre-and Peri-Natal Psychology Journal, 11(2), 81.
Postnatal depression in New Zealand
31
McGill, H., Burrows, V., Holland, L., Langer, H., & Sweet, M. (1995). Postnatal depression: a
Christchurch study. The New Zealand Medical Journal, 108(999), 162–165.
Ministry of Health. (2004). Ethnicity data protocols for the health and disability sector.
Thio, I. M., Browne, M. A. O., Coverdale, J. H., & Argyle, N. (2006). Postnatal depressive
symptoms go largely untreated: A probability study in urban New Zealand. Social
Psychiatry and Psychiatric Epidemiology, 41(10), 814–818. http://doi.org/10.1007/s00127006-0095-6
Waldie, K. E., Peterson, E. R., D’Souza, S., Underwood, L., Pryor, J. E., Carr, P. A., … Morton, S.
M. B. (2015). Depression symptoms during pregnancy: Evidence from Growing Up in New
Zealand. Journal of Affective Disorders, 186, 66–73.
http://doi.org/10.1016/j.jad.2015.06.009
Webster, M. L., Thompson, J. M., Mitchell, E. A., & Werry, J. S. (1994). Postnatal depression in a
community cohort. Australian and New Zealand Journal of Psychiatry, 28(1), 42–49.
32
Postnatal depression in New Zealand